1. Field of the Invention
The present invention relates to a technique for describing a multiple level digest segment information scheme for multimedia contents in a content-based data area of a multimedia stream and method for generating a digest streams(highlights) using the multiple level digest segment information scheme for the purpose of providing a multiple levels of digest streams for a multimedia stream.
2. Description of the Prior Art
A digest stream for a multimedia stream is provided for the purpose of enabling users to understand the overall story and the structure of a multimedia content or to find and move to desired position fast.
Multimedia server system such as a VOD(Video-on-Demand) server system, a broadcast system, a karaoke, etc., can provide digest streams(highlights) for multimedia contents (e.g., movie, drama, sports, etc.) to achieve above purposes.
A digest stream(highlight) of a multimedia stream is constructed by arranging audio-visual segments, each of which is the essence of an original stream in the development of the story. Here, the segment is a sequence of continuous frames. Since the digest stream is a sub stream of the multimedia stream, the size of the digest stream is much smaller than that of the original stream.
Therefore, a user who want to understand the overall structure and/or the story of a multimedia content prefers to access a digest stream rather than accessing the entire stream. The digest stream also can be used as a program guide for users who want to select a program from a multimedia database.
There are two types of summarization method for multimedia contents. The one is providing key frames or key regions and the other is providing digest streams(highlights) which is constructed by gathering and arranging important segments of an original stream
In the former method, since it is difficult for a user to understand the entire flow of the original stream and it is difficult to display too many key frames at a time, an additional operation such as a screen scrolling is required in order to understand the content of the stream only with the key frames, thereby causing an inconvenience.
In the latter method, generally, a content provider extracts a sub stream from the original stream, and stores it separately. In this case, an additional storage for storing the digest stream is required. In addition, the content provider edits only a certain time amount of the digest stream and provides it to the user. Accordingly, the user can access only a single level digest stream edited by the content provider. Therefore, users cannot obtain various digest streams of the desired length. Suppose that a content provider provides only a 10 minute highlight for a 2 hour video. Under that situation, users may want to access 5, 10, 20 minute digest streams respectively, depending upon their situation. In that case, a user who want 5 or 20 minute highlight for the given content can not browse the digest stream of their desire. Only 10 minute digest stream will be served regardless of the user's request. Therefore, in order to satisfy user's various needs, the content provider has to edit and store multiple levels of digest streams for one multimedia stream. In this case, a large amount of storage is required to store multiple versions of digest streams, and accordingly there occurs difficulty in managing each of the digest stream.